Colchicine modulates clonal hematopoiesis–associated inflammatory signaling in coronary artery disease: the COLCHIP trial

Background/Introduction

Clonal haematopoiesis of indeterminate potential (CHIP) promotes pro-inflammatory signalling through enhanced nuclear factor kappa-B (NF-κB) and inflammasome activation, contributing to residual risk in coronary artery disease (CAD). Whether CHIP-associated inflammatory networks in humans are dynamically and pharmacologically modifiable remains unknown.

Purpose

To determine whether colchicine disrupts clonal burden–linked inflammatory proteomic networks in stable CAD.

Method

COLCHIP was a randomised, double-blind, placebo-controlled crossover trial in patients with angiographically documented stable CAD and CHIP (variant allele frequency ≥2%). Participants received colchicine 0.5 mg once daily or placebo for 4 weeks, followed by crossover. Plasma proteomics (5,416 proteins) were quantified at baseline (T1), during colchicine exposure (T2), and 4 weeks after cessation (T3). Linear mixed-effects models assessed treatment effects (false-discovery rate q<0.05). Associations between summed clonal burden and protein abundance were evaluated by correlation analysis and pathway enrichment.

Results

54 patients were randomised; 48 were analysed per protocol (mean age 70±6 years; 81% male). Median summed variant allele frequency was 5.2% (interquartile range 2.7–12.1).

At baseline, clonal burden was tightly coupled to coordinated inflammatory protein networks enriched in NF-κB, tumour necrosis factor, leukocyte adhesion, and mitogen-activated protein kinase pathways (all enrichment P<0.001), defining a CHIP-associated inflammatory endotype.

Colchicine exposure reduced leukocyte counts (P=0.02) but did not modify C-reactive protein or interleukin-6. In contrast, high-dimensional proteomics revealed selective network remodelling: five proteins were differentially expressed at T2 (q<0.05), and 169 at T3, predominantly mapping to NF-κB and leukocyte activation pathways. Importantly, baseline correlations between clonal burden and inflammatory protein modules were markedly attenuated during colchicine exposure, particularly within mitogen-activated protein kinase–enriched clusters, with partial re-coupling after cessation. No treatment-by-period interaction was observed.

Conclusion

In stable CAD with CHIP, colchicine disrupts clonal burden–coupled inflammatory signalling at the network level without affecting conventional biomarkers. These data demonstrate that CHIP defines a pharmacologically modifiable inflammatory endotype characterised by coordinated NF-κB– and mitogen-activated protein kinase–enriched signalling. Proteomic network remodelling provides a mechanistic framework for genotype-stratified anti-inflammatory intervention.